SECOND GENERATION ADVANCED REBURNING FOR HIGH EFFICIENCY NO x

Transcription

1 SECOND GENERATION ADVANCED REBURNING FOR HIGH EFFICIENCY NO x CONTROL DOE Contract No. DE-AC22-95PC95251 Final Report Project Period: October, June, 2001 Prepared by: Vladimir M. Zamansky, Peter M. Maly, Vitali V. Lissianski, Mark S. Sheldon, David Moyeda, and Roy Payne Submitted by: GE Energy and Environmental Research Corporation 18 Mason, Irvine, CA June 27, 2001

2 g DOE Contract No. DE-AC22-95PC95251 Final Report Disclaimer This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. i

3 g DOE Contract No. DE-AC22-95PC95251 Final Report Abstract This project develops a family of novel Second Generation Advanced Reburning (SGAR) NO x control technologies, which can achieve 95% NO x control in coal fired boilers at a significantly lower cost than Selective Catalytic Reduction (SCR). The conventional Advanced Reburning (AR) process integrates basic reburning and N-agent injection. The SGAR systems include six AR variants: (1) AR-Lean - injection of the N-agent and promoter along with overfire air; (2) AR-Rich - injection of N-agent and promoter into the reburning zone; (3) Multiple Injection Advanced Reburning (MIAR) - injection of N-agents and promoters both into the reburning zone and with overfire air; (4) AR-Lean + Promoted SNCR - injection of N-agents and promoters with overfire air and into the temperature zone at which Selective Non-Catalytic Reduction (SNCR) is effective; (5) AR-Rich + Promoted SNCR - injection of N-agents and promoters into the reburning zone and into the SNCR zone; and (6) Promoted Reburning + Promoted SNCR - basic or promoted reburning followed by basic or promoted SNCR process. The project was conducted in two phases over a five-year period. The work included a combination of analytical and experimental studies to confirm the process mechanisms, identify optimum process configurations, and develop a design methodology for full-scale applications. Phase I was conducted from October, 1995 to September, 1997 and included both analytical studies and tests in bench and pilot-scale test rigs. Phase I moved AR technology to Maturity Level III - Major Subsystems. Phase II is conducted over a 45 month period (October, 1997 June, 2001). Phase II included evaluation of alternative promoters, development of alternative reburning fuel and N-Agent jet mixing systems, and scale up. The goal of Phase II was to move the technology to Maturity Level IV- Subscale Integrated System. Tests in combustion facility ranging in firing rate from to Btu/hr demonstrated the viability of the AR technology. The performance goals of the project to reduce NO x by up to 95% with net emissions less than 0.06 lb/10 6 Btu and to minimize other pollutants (N 2 O and NH 3 ) to levels lower than reburning and SNCR have been met. Experimental data demonstrated that AR- Lean + SNCR and Reburning + SNCR are the most effective AR configurations, followed by AR- Lean and AR-Rich. Promoters can increase AR NO x reduction efficiency. Promoters are the most effective at small amounts of the reburning fuel (6-10% of the total fuel heat input). Promoters provide the means to improve NO x reduction and simultaneously decrease the amount of reburning fuel. Tests also showed that alkali-containing compounds are effective promoters of the AR process. When co-injected with N-agent, they provide up to 25 % improvement in NO x reduction. A detailed reaction mechanism and simplified representation of mixing were used in modeling of AR processes. Modeling results demonstrated that the model correctly described a wide range of experimental data. Mixing and thermal parameters in the model can be adjusted depending on characteristics of the combustion facility. Application of the model to the optimization of AR-Lean has been demonstrated. Economic analysis demonstrated a considerable economic advantage of AR technologies in comparison with existing commercial NO x control techniques, such as basic reburning, SNCR, and SCR. Particularly for deep NO x control, coal-based AR technologies are 50% less expansive than SCR for the same level of NO x control. The market for AR technologies is estimated to be above $110 million. ii

4 g DOE Contract No. DE-AC22-95PC95251 Final Report Table of Contents Section Page Abstract... ii Table of Contents... iii List of Figures... vii List of Tables... xiii Nomenclature... xiv Executive Summary... xvi 1.0 Introduction Background High Efficiency NO x Control under Title 1 of the CAAA Limitations of Available NO x Control Technologies for Post-RACT Applications Advanced Reburning Second Generation Advanced Reburning (SGAR) Phase I Results Phase I Program Objectives Summary of Phase I Results Phase I Conclusions Phase II Program Approach, Objectives and Tasks Phase II Technical Approach Phase II Objectives and Tasks Task 2.1 Project Coordination and Reporting/Deliverables Task 2.2 Studies of Prospective Additives Bench-Scale Screening Tests Pilot-Scale Test Results Effect of Metals on NO x Reduction Effect of Ca on SO 2 Emissions Effect of Iron-Containing Compounds on NO x Reduction Effects of Fly Ash and Char on NO x Reduction Prospective Additives Search: Summary Task 2.3 Development of Combined Chemistry/Mixing Model iii

5 g DOE Contract No. DE-AC22-95PC95251 Final Report 7.1 Model Setup Model Formulation Estimation of Mixing Parameters Mixing Mode Mixture Stratification Chemistry Mixing Modeling of Gas Reburning Comparison with Experimental Data Obtained in Bench- and Pilot-Scale Facilities Comparison with Experimental Data of Kolb et al Comparison with Experimental Data of Mereb and Wendt Parametric Study of Basic Reburning Effect of Fuel Stratification in the Reburning Zone Effect of the Initial Temperature of the Reburning Fuel and Overfire Air Reactions Responsible for NO x Reduction Gas Reburning Modeling: Summary Chemistry-Mixing Modeling of Ammonia and Sodium Effects on Reburning Combined Injection of N-Agent and Sodium Promoter Injection of Promoters without N-Agent Chemistry-Mixing Modeling of Ammonia and Sodium Effects: Summary Optimization of AR via Modeling AR-Lean Model Setup Model Validation Parametric Study of the AR-Lean Process Mapping of the AR Process AR Optimization: Summary Task 2.4 Optimization of Process Synergism in Btu/hr Tests Pilot-Scale Optimization Tests Description of Greenidge AR System Simulation of Greenidge Boiler Baseline and Gas Reburning NO x Data Effect of Pulsing on Basic Reburning iv

6 g DOE Contract No. DE-AC22-95PC95251 Final Report AR-Lean Test Results Comparison between Greenidge and Pilot-Scale AR-Lean Data Reburning + SNCR Comparison of Greenidge and Pilot-Scale Data: Summary Coal Reburning Studies Basic Coal Reburning Tests Advanced Coal Reburning Tests Advanced Coal Reburning: Summary Task Btu/hr Proof-of-Concept Tests AR-Lean Tests AR-Rich Tests Reburning + SNCR Tests MIAR Test Results Ammonia Slip and N 2 O Emissions Measurements Proof-of-Concept Tests: Summary Task 2.6 Design Methodology Validation Full-Scale Implementation of AR Activities at Martinez Refining Company Complex AR-Lean Tests in Large Pilot-Scale AR Economic and Market Update NO x Control Drivers Economic Methodology and Case Studies Technology Specific Inputs Economic Results Market Assessment Economic and Market Analysis: Conclusions Design Methodology and Application Conclusions Experimental Facilities v

7 g DOE Contract No. DE-AC22-95PC95251 Final Report 11.1 Process Performance Characterization Controlled Temperature Tower (CTT) Boiler Simulator Facility (BSF) Tower Furnace (TF) Conclusions References Appendix A Phase I Report... A-1 Appendix B Reaction Mechanism of C-H-O-N Species in Chemkin Format... B-1 Appendix C Thermodynamic Database for C-H-O-N Species in Chemkin Format... C-1 Appendix D Reaction Mechanism of Na Species in Chemkin Format... D-1 Appendix E Thermodynamic Database for Na Species in Chemkin Format... E-1 Appendix F Spray Evaporation Modeling Studies... F-1 vi

8 g DOE Contract No. DE-AC22-95PC95251 Final Report List of Figures Figure Page Figure 1-1. Schematic of different variants of AR Figure 5-1. Phase II task structure Figure 6-1. Alternative promoter AR-Rich screening test results Figure 6-2. Performance of alternative promoter performance as a Function of injection temperature Figure 6-3. Effect of FeSO 4 on the reburning process Figure 6-4. Promoter screening tests Figure 6-5. Effect of metal-containing compounds injected with the main fuel on NO x reduction Figure 6-6. Na and K performance as a function of promoter concentration Figure 6-7. NO x reduction as function of Na concentration Figure 6-8. Injection of 100 ppm Na and Ca into reburning zone at 1590 K Figure 6-9. Effect of 100 ppm Na addition on CO emissions Figure Calcium promoter NO x control performance during Utah coal firing Figure Calcium promoter SO 2 capture during Utah coal firing Figure Test data on NO x reduction in the presence of iron-containing compounds Figure Effect of fly ash co-injection with reburning fuel on NO x reduction Figure Reburning performance of activated char as a function of char reburning heat input Figure 7-1. Reactor diagram of model setup Figure 7-2. A diagram of jet injection and model setup in the BSF reburning zone Figure 7-3. Modeling setup in the BSF mixing zone Figure 7-4. Comparison of BSF test results on reburning with modeling predictions Figure 7-5. Performance of basic reburning in CTT Figure 7-6. Performance of basic reburning in TF Figure 7-7. Predicted effect of mixing time on NO x reduction for typical BSF conditions Figure 7-8. Modeling and experimental data on concentrations of N-containing species at the end of the BSF reburning zone Figure 7-9. Comparison of experimental results on basic reburning with modeling predictions vii

9 g DOE Contract No. DE-AC22-95PC95251 Final Report Figure Comparison of modeling predictions with experimental data of Kolb et al Figure Effect of the reburning zone stoichiometry on efficiency of the reburning process Figure Comparison of modeling predictions with experimental data of Mereb and Wendt Figure Effect of fuel stratification in the mixing area of the reburning zone on modeling predictions for BSF conditions Figure Predicted effect of the initial temperature of the reburning fuel on NO reduction for BSF conditions Figure Predicted effect of the initial OFA temperature on NO reduction for BSF conditions Figure Performance of the reburning process for optimized and non optimized initial temperatures of the injected reburning fuel and overfire air Figure Concentrations of N-containing species in reburning and main reactions responsible for NO x reduction in different zones Figure Comparison of modeling predictions with experimental data Figure Rate coefficient of the reaction NaOH + H Na + H 2 O Figure Comparison of modeling predictions with experimental data on CO emissions Figure CO emissions in the AR process as a function of flue gas temperature at the point of OFA injection Figure Comparison of modeling predictions with experimental data on the effect of Na 2 CO 3 injection in BSF at 18% reburning Figure Reactor diagram of AR-Lean model setup Figure Comparison of modeling predictions with experimental data on the effect of OFA/urea injection temperature on NO x reduction in AR-Lean at 2% and 10% reburning Figure Comparison of modeling predictions with experimental data on basic reburning and AR-Lean reburning Figure Effect of CO on NO x reduction by urea injection Figure Predicted effect of the reburning heat input on NO x reduction at different OFA injection temperatures viii

10 g DOE Contract No. DE-AC22-95PC95251 Final Report Figure Predicted effect of OFA/urea injection temperature in AR-Lean Figure Effect of droplet evaporation time on NO x reduction Figure Predicted effects of droplet evaporation time on NO x reduction at 10% and 5% reburning Figure Predicted effect of the amount of urea on NO x reduction Figure Predicted effects of OFA and urea initial temperature on NO x reduction Figure Performance of the AR-Lean process at NSR= Figure Performance of the AR-Lean process at NSR= Figure Performance of AR-Lean at NSR=0.7 as a function of the amount of the reburning fuel and droplet evaporation time of N-agent Figure Predicted performances of basic reburning, AR-Lean and Reburning+SNCR Figure 8-1. Overview of the advanced gas reburning system installed on the Greenidge boiler Figure 8-2. Comparison of Greenidge 1996 and 1997 gas reburning data Figure 8-3. Comparison of gas reburning performance between Greenidge and BSF Figure 8-4. Effect of fluctuations on reburning Figure 8-5. AR-Lean performance vs. NSR at different CO concentrations Figure 8-6. AR-Lean performance vs. NSR at different CO concentrations Figure 8-7. AR-Lean performance vs. reburning zone CO concentration at 5% and 10% reburning Figure 8-8. Incremental performance of N-agent alone in AR-Lean as a function of reburning zone CO concentration Figure 8-9. AR-Lean performance versus additive injection temperature with and without main fuel pulsing at 10% reburning Figure AR-Lean performance versus additive injection temperature with main fuel pulsing at 5% reburning Figure AR-Lean performance versus NSR at initial NO concentrations of 300 and 600 ppm at 10% reburning Figure NO reduction in AR-Lean as a function of SR 2 with and without pulsing Figure Comparison of the impacts of CO concentration on NO x reduction in AR-Lean tests in Greenidge and BSF ix

11 g DOE Contract No. DE-AC22-95PC95251 Final Report Figure Comparison of the impacts of injection temperature on NO x reduction in Greenidge and BSF Figure NO x reduction due to ammonia in AR-Lean in Greenidge and BSF Figure SNCR performance as a function of N-agent injection temperature with and without main fuel pulsing Figure Comparison of performance of different N-agents in the reburning + SNCR process without pulsing Figure Comparison of performance of different N-agents as a function of NSR at 1270 K and1300 K (without pulsing) Figure Comparison of results obtained with the three N-agents during pulsing of the main fuel Figure NO reduction in reburning + SNCR as a function of pulsing frequency at 10% and 20% reburning Figure Comparison of results obtained at pulsing amplitudes of 5% and 10% Figure Promoted reburning + SNCR performance as a function of additive injection temperature at 10% reburning Figure Promoted reburning + SNCR performance as a function of sodium concentration at 10% reburning Figure Promoted reburning + SNCR performance as a function of reburning heat input Figure Ammonia slip results at different reburning + SNCR process conditions Figure Basic coal reburning performance as a function of reburning heat input with natural gas primary Figure Basic coal reburning performance as a function of reburning heat input with coal primary Figure AR-Lean performance at 10% reburning as a function of injection temperature for coals #1 and # Figure Promoted AR-Lean performance at 10% reburning as a function of promoter concentration for coals #1 and # Figure AR-Rich performance at 10% reburning as a function of injection temperature for coals #1 and # x

12 g DOE Contract No. DE-AC22-95PC95251 Final Report Figure Promoted AR-Rich performance at 10% reburning as a function of promoter concentration for coals #1 and # Figure Combined reburning and SNCR performance at 10% reburning as a function of injection temperature for coals #1 and # Figure Combined reburning and SNCR performance as a function of reburning heat input for coals #1 and # Figure Combined reburning and SNCR performance at 15% reburning as a function of promoter concentration for coals #1 and # Figure Effect of Fe promoter concentration on reburning Figure 9-1 AR-Lean performance vs. promoter concentration at 6% and 10 % reburning Figure 9-2. AR-Rich performance vs. promoter concentration at 6% and 10 % reburning Figure 9-3. Reburning+SNCR performance vs. promoter concentration at 10% and 20% reburning Figure 9-4. MIAR: combined AR-Lean + AR-Rich performance vs. promoter concentration at 10% reburning Figure 9-5. MIAR: combined AR-Lean + SNCR performance vs. promoter concentration at 6% and 10% reburning Figure 9-6. Overall MIAR NO x reduction under optimized conditions at 6% reburning Figure Atomization characteristics of test nozzle Figure Temperature profiles for the target boiler and the TF Figure Impact of atomization pressure for 5% urea solution on overall (a) and incremental (b) NO x reduction by urea solution co-injected with OFA Figure Incremental NO x reduction for co-injection of 5% urea solution with OFA Figure Incremental NO x reduction for co-injection of different urea solution strengths with OFA Figure Comparison of overall NO x reduction by urea solution for 3-port and 4-port configurations of OFA injection Figure Overall NO x reduction at different urea injector positions. NSR = Figure Comparison of performance for urea and aqueous ammonia as a function of N-agent atomizing air pressure Figure Comparison of overall NO x reduction at different OFA/urea injection temperatures xi

13 g DOE Contract No. DE-AC22-95PC95251 Final Report Figure SNCR performance as a function of atomization pressure with and without reburning Figure Axial injector tests: impact of atomization pressure on overall NO x reduction at different reburning heat inputs Figure Axial injector tests: comparison of urea and ammonium sulfate impacts on NO x reduction Figure Comparison of cyclone-fired boiler NO x control technology economics Figure Comparison of wall-fired boiler NO x control technology economics Figure Impact of fuel differential on cyclone-fired boiler NO x control economics Figure Impact of fuel differential on wall-fired boiler NO x control economics Figure Axial temperature profiles measured in CTT, BSF and TF Figure Controlled Temperature Tower (CTT) Figure CTT temperature profiles Figure Boiler Simulator Facility (BSF) Figure Schematic diagram of the Tower Furnace (TF) xii

14 g DOE Contract No. DE-AC22-95PC95251 Final Report List of Tables Table Page Table 2-1. Performance of NO x control technologies Table 2-2. AR variants Table 6-1. Compounds tested in CTT as advanced reburning promoters Table 6-2. Mineral composition of fly ash generated by combustion of a Kentucky coal Table 7-1. Characteristics of mixing in BSF and jet parameters Table 7-2. Mixing parameters in CTT, BSF and TF Table 7-3. Reburning parameters Table 8-1. Test fuel analyses Table 9-1. Results of ammonia slip tests Table Economic data Table NO x control technologies and expected performance Table NO x control technology economics Table Comparing the cost effectiveness for deep NO x control xiii

15 g DOE Contract No. DE-AC22-95PC95251 Final Report Nomenclature AR - Advanced Reburning AR-Lean - Advanced Reburning Lean AR-Rich - Advanced Reburning Rich BSF - Boiler Simulator Facility Btu - British thermo units CAAA - Clean Air Act Amendment CFD - Computational Fluid Dynamics CRF - Capital Recovery Factor CTT - Controlled Temperature Tower DOE - Department of Energy EPRI - Electric Power Research Institute ESP - Electrostatic Precipitator FETC - Federal Energy Technology Center (U.S. Department of Energy) GC - Gas Chromatography GE-EER - General Electric Energy & Environmental Research Corp. GRI - Gas Research Institute JICFIS - Jets In Crossflow, Integral Solution code LNB - Low NO x Burners MIAR - Multiple Injection Advanced Reburning MS - Mass-Spectrometry NAAQS - National Ambient Air Quality Standards NEOTR - Northeast Ozone Transport Region NESCAUM - Northeast States for Coordinated Air Use Management NETL - National Energy Technology Laboratory (U.S. Department of Energy) NSR - Nitrogen Stoichiometric Ratio ODF - One Dimensional Flame code OFA - Overfire Air PFR - Plug Flow Reactor RACT - Reasonably Available Control Technologies SCR - Selective Catalytic Reduction SGAR - Second Generation Advanced Reburning SIP - Strategic Implementation Plan xiv

16 g DOE Contract No. DE-AC22-95PC95251 Final Report SNCR - Selective Non-Catalytic Reduction TAG - Technology Assessment Guide TF - Tower Furnace TFN - Total Fixed Nitrogen WSR - Well-Stirred Reactor xv

17 g DOE Contract No. DE-AC22-95PC95251 Final Report Executive Summary This project develops a family of novel Second Generation Advanced Reburning (SGAR) NO x control technologies, which have the potential to achieve 95% NO x control in coal fired boilers at a significantly lower cost than Selective Catalytic Reduction (SCR). AR systems integrate basic reburning and injection of an N-agent (a nitrogen-containing species, typically ammonia or urea, capable of converting NO to N 2 ). Specific features of the new AR systems in comparison with basic reburning include: Introduction of reburning fuel representing a small portion of the total fuel heat input, to provide slightly fuel-rich conditions in the reburning zone. N-agent injection at one or two locations, which may include the reburning zone, the point of overfire air injection, and/or downstream of overfire air injection. Injection of promoter additives which enhance the effectiveness of the N-agent. The Advanced Reburning (AR) process is a GE-EER patented synergetic integration of basic reburning and N-agent injection. In this process, an N-agent is injected along with the overfire air (OFA) and the reburning system is adjusted to optimize the NO x reduction due to the N-agent. By adjusting the reburning fuel injection rate to achieve near stoichiometric conditions (instead of the fuel rich conditions normally used for reburning), the CO level is controlled and the temperature window for selective NO x reduction is broadened and deepened. The reburning fuel requirement is reduced from about 20% of total fuel heat input for basic reburning, to about 10% for AR, which has considerable economic benefits (the incremental cost of gas for gas reburning and the cost of the coal pulverization equipment for coal reburning). With AR, the NO x control due to reburning fuel addition is somewhat reduced from basic reburning; however, this reduction is offset by the significant enhancement of the N-agent NO x control. This project develops AR systems which broaden technology applicability to a wide range of boiler designs. The AR systems incorporate several improvements over conventional AR, such as N-agent injection into the reburning zone, promoter additives which enhance the effectiveness of the N- agent, and injection of N-agents with or without promoters at two locations. Sodium salts, in particular sodium carbonate (Na 2 CO 3 ), were identified as effective AR promoters. Salts of other alkali metals can also be used as promoters. This family of AR technologies is intended for post- RACT applications in ozone non-attainment areas where NO x control in excess of 70%-80% is required. The AR systems are applicable to all types of coal fired boilers without massive hardware changes, without increasing air toxic and toxic waste problems, and at a cost for NO x control on the order of half that of SCR. These systems will provide flexible installations and do not create secondary pollutants and can be integrated with SO 2 and air toxics control methods. They are also highly flexible, in that components can be added over time as NO x emissions regulations become more stringent. Selection of a technology for a specific boiler can be made based on boiler access, thermal conditions, and NO x control requirements. In the AR processes, the N-agent can be injected with or without promoters at one or two of three chemically significant locations: into the reburning zone, along with OFA, or downstream of burnout in the temperature window for which Selective Non-Catalytic Reduction (SNCR) is effective, the SNCR zone. Accordingly, there are six AR variants: xvi

18 g DOE Contract No. DE-AC22-95PC95251 Final Report Promoted Advanced Reburning Lean (AR-Lean): Injection of the N-agent and promoter along with overfire air. Promoted Advanced Reburning Rich (AR-Rich): Injection of N-agent and promoter into the reburning zone. Multiple Injection Advanced Reburning (MIAR): Injection of N-agents and promoters both into the reburning zone and with overfire air. AR-Lean + Promoted SNCR: Injection of N-agents and promoters with overfire air and into the SNCR zone. AR-Rich + Promoted SNCR: Injection of N-agents and promoters into the reburning zone and into the SNCR zone. Reburning + Promoted SNCR: Basic or promoted reburning followed by basic or promoted SNCR process. In each of these variants, the use of promoters is optional. When employed, promoters are typically co-injected with the N-agent. The project was conducted in two phases over a five-year period. The work included a combination of analytical and experimental studies to confirm the process mechanisms, identify optimum process configurations, and develop a design methodology for full-scale applications. Phase I was conducted from October, 1995 to September, 1997 and included both analytical studies and tests in bench and pilot scale test rigs. Phase I moved AR technology to Maturity Level III - Major Subsystems. Phase II is conducted over a 45 month period (October, 1997 June, 2001). Phase II is built on the Phase I results and includes evaluation of alternate promoters, development of alternative reburning fuel and N-Agent jet mixing systems, and scale-up. The goal of Phase II was to move the technology to Maturity Level IV- Subscale Integrated System. The overall objective of Phase I was to demonstrate the effectiveness of the AR technologies at bench and pilot scale over a sufficiently broad range of conditions to provide all of the information needed for process optimization and scale up. Specific program objectives were as follows: 1. Develop an understanding of the mechanisms through which promoter additives improve N- agent effectiveness; 2. Develop a kinetic analytical model of the Promoted and Multiple Injection AR technologies; 3. Optimize the AR processes using the analytical model and the results of bench and pilot scale experiments under controlled mixing conditions; and 4. Upgrade GE-EER s AR design methodology to accommodate the technical advancements of AR. Phase I project determined the ability of the AR technologies to meet the following technical performance goals: NO x emissions from the Btu/hr coal fired Boiler Simulator Facility controlled to less than the requirements for post-ract NO x control in the NESCAUM area for the year 2003; Total estimated cost of controlling NO x emissions, based on the Btu/hr coal fired tests, shown to be less than that currently projected for SCR NO x control systems; and No significant reduction in boiler efficiency or significant adverse environmental impacts when compared to current reburning and SNCR technologies. xvii

19 g DOE Contract No. DE-AC22-95PC95251 Final Report Phase I consisted of the following six tasks: Task 1.1 Task 1.2 Task 1.3 Task 1.4 Task 1.5 Task 1.6 Project coordination and reporting/deliverables. Kinetics of Na 2 CO 3 reactions with flue gas components Btu/hr optimization studies Btu/hr process development tests. Mechanism development and modeling. Design methodology and application. A flow system decomposition study in Task 1.2 revealed that the primary gas-phase decomposition products of Na 2 CO 3 are Na atoms, NaOH and CO 2. Extrapolating the results to higher temperatures showed that Na 2 CO 3 decomposition at temperatures over 1400 K produced NaOH and CO 2 very quickly. NaOH then decomposed more slowly. These findings were incorporated into kinetic modeling in Task 1.5. In Tasks 1.3 and 1.4 bench scale combustion tests in the Btu/hr facility were conducted. These tests demonstrated NO x reduction of 86%, 88%, and 91% for AR-Lean, AR-Rich, and MIAR, respectively. These levels of NO x control were achieved with only 15 ppm Na 2 CO 3 in flue gas. Pilot scale studies in the Btu/hr combustion facility demonstrated the ability of the AR technologies to achieve NO x reductions of 95+% during gas firing and 90+% during coal firing. Byproduct emissions were found to be lower than those generated by commercial reburning and SNCR technologies. In Task 1.5 a detailed reaction mechanism was developed to model the AR chemical processes. Kinetic modeling provided insight into the controlling factors of the process and qualitatively described the observed reaction trends. Modeling predicted that the following factors mainly defined the efficiency of AR systems: equivalence ratio in the reburning zone, process streams injection temperatures (reburning fuel, N-agents, promoters, and OFA), concentrations of N-agents and promoters, delay times for injection of N-agents into the reburning and burnout zones, and characteristic mixing times of the injection streams with flue gas. The modeling predicted and explained the NO x reduction enhancement of sodium promotion under both fuel-rich and fuel-lean conditions. The AR design methodology was upgraded in Task 1.6 using experiments and analytical models to include the second generation improvements. This work took advantage of a full-scale demonstration of the original AR technology, already in progress under separate project funding, on a 105 MW tangentially fired boiler. The upgraded methodology was used to prepare process designs for three AR concepts on the 105 MW boiler, and to predict the impacts of the AR systems on boiler performance and NO x emissions. Some elements of AR were tested in the boiler. These tests showed that the large scale stratification in the furnace gases affected the NO x reduction and ammonia slip associated with N-agent injection. An economic analysis was conducted to compare the cost effectiveness of AR and SCR using the EPRI Technology Assessment Guide methodology for two representative Title 1 CAAA applications: a cyclone fired boiler and a wall fired boiler equipped with low NO x burners. The total cost of NO x control (combining capital and operating cost components) for the AR systems was 48-69% less than for SCR depending on the specific application. The requirements for NO x control under the CAAA were evaluated. The key drivers to implement AR are the current ozone nonattainment areas, the potential to expand those regions to the eastern half of the U.S., and the recent xviii

20 g DOE Contract No. DE-AC22-95PC95251 Final Report tightening of the National Ambient Air Quality Standards for ozone and fine particulate which will require additional NO x control nationwide. The market for AR technologies was estimated to be above $1.5 billion. Phase II filled the gap between the Phase I development and a long-term AR demonstration by doing the following: Identify alternative promoters based on the promotion mechanisms developed in Phase I. Identify and test coal mineral compounds responsible for the increased NO x reduction in AR-Rich and MIAR with coal firing (about 10% higher than for gas firing). Optimize mixing (of reburning fuel, N-agents, and OFA into the furnace gas stream) via combined chemistry/mixing models. Optimize N-agent injection to maximize NO x reduction with negligible ammonia slip. Evaluate the effect of N-agent/promoter mixing times representative of full scale. Optimize AR with new promoters and mixing regimes at Btu/hr scale. Scale up and confirm the design methodology via Btu/hr Proof-of-Concept tests as well as limited component tests conducted during the ongoing boiler AR tests. Update the economic and market analysis to confirm the advantages of AR. Specific Phase II objectives were to: 1. Develop alternative NO x control promoters for AR. 2. Develop a combined chemistry/mixing model of the process to optimize mixing regimes. 3. Confirm the design methodology via pilot scale experiments at and Btu/hr. Phase II also determined the ability of the AR technologies to meet the following technical performance goals in the Btu/hr Proof-of-Concept coal firing tests: 1. Reduce NO x by 95% with net emissions less than 0.06 lb NO x /10 6 Btu. 2. Minimize other pollutants (N 2 O and NH 3 ) to levels lower than reburning and SNCR. 3. Minimize net parasitic power consumption to less than 0.5% of the power plant energy. 4. Minimize the total cost of NO x control to less than half that of SCR. Phase II included the following tasks: 2.1 Project coordination and reporting/deliverables. 2.2 Studies of other prospective promoters. 2.3 Development of a combined chemistry/mixing model. 2.4 Optimization of process synergism in Btu/hr tests Btu/hr proof-of-concept tests. 2.6 Design methodology validation. In Task 2.2 the effects of additives on AR-Rich and basic reburning were determined. Tests showed that co-injection of Li and K compounds resulted in 74-78% NO reduction, i.e percentage points improvement above the baseline reburning level. Although these effects are lower than those for sodium, they are significant. Thus, K and Li compounds can be used as AR promoters. xix

21 g DOE Contract No. DE-AC22-95PC95251 Final Report Compounds of Mg, Ca, Ba, and Zn provided relatively small promotional effect. When added to ammonia solution, they reduced NO by an additional 6-9 percentage points compared to unpromoted AR. Tests also showed that metal-containing compounds could be effective reburning promoters without injection of N-agent. Fe-containing compounds were the most effective in reduction of NO x emissions, followed by Na-, K-, and Ca-containing compounds. Co-injection of these compounds with the main fuel in the absence of reburning resulted in 16-30% NO x reduction. Injection of metal compounds with the main fuel in the presence of reburning provided an additional 4-25% percentage points of NO x reduction above the baseline reburning level. As the concentration of additive increased, so did the promotional effect. Co-injection of additives with reburning fuel and into the reburning zone had smaller effect than co-injection with the main fuel. Coal char and fly ash showed minimal effect on NO x reduction. It is concluded that metals in coal char and fly ash were mainly present in the form of sulfides and silicate-alumosilicate matrixes that were more stable than carbonates and acetates at high temperatures. These compounds were not effective in reactions with combustion radicals and have a minimal effect on NO x reduction. Tests showed that not only did injection of Ca-containing compounds reduce NO x emissions, but it also decreased SO 2 emissions: about 50% SO 2 reduction was achieved with the injection of 1,000 ppm of Ca(OH) 2 with main fuel. The model of AR processes was updated in Task 2.3. Modeling results demonstrated that the model correctly described a wide range of experimental data obtained in five bench- and pilotscale combustion facilities. This suggested that the model, as developed through Phase II, represented the main chemical and mixing features of the reburning process and could be used for process optimization. Mixing and thermal parameters in the model can be adjusted depending on the characteristics of the combustion facility. The following conclusions were drawn from modeling results: Stratification in the mixing zone improves reburning efficiency for small heat inputs of the reburning fuel and degrades reburning efficiency for large heat inputs. Based on modeling observations, it is suggested that design of the nozzle for the reburning fuel injection should be different depending on the amount of the injected reburning fuel. Injection of large amounts of the reburning fuel provides better NO x reduction if mixing of reactants is fast. Injection of small amounts of the reburning fuel, on the other hand, should result in significant mixture stratification for better NO x control (as long as complete mixing and burnout is ultimately achieved). Initial temperatures of the reburning fuel and OFA affect NO reduction and can be optimized for deeper NO control. Optimum temperatures depend on the mixture composition and on the injection location. By optimizing these parameters, NO x reduction can be increased by several percentage points. Reactions of NH 3 in the burnout zone play an important role in NO reduction for large heat inputs of the reburning fuel. The applicability of the model to the optimization of AR-Lean has been demonstrated. Modeling identified the following AR-Lean parameters as being most important: amounts of the reburning fuel and N-agent, temperature of flue gas at the point of OFA/N-agent injection, and evaporation time of the N-agent. Modeling predictions, supported by experiments, are that CO formed in the xx

22 g DOE Contract No. DE-AC22-95PC95251 Final Report reburning zone increases the efficiency of N-agent when the temperature of furnace gases at the point of OFA/N-agent injection is lower than 1200 K, and reduces its efficiency at higher injection temperatures. To reduce the negative effect of CO on NO x reduction at OFA/N-agent injection temperatures typically utilized in utility boilers, the average droplet size of injected N-agent solution must be optimized to allow for CO oxidation in the burnout zone before a significant amount of N- agent evaporates. In Task 2.4 BSF tests were conducted to determine the optimum process conditions at mixing and thermo characteristics of Greenidge 105 MW tangentially fired boiler. Tests focused on simulating the AR-Lean and reburning + SNCR performance as the most promising AR variants for deep NO x control for the Greenidge unit. The results of the BSF simulation tests demonstrated that high CO concentrations typical for upper furnace of the Greenidge boiler have negative effects on AR-Lean performance at the NH 3 /OFA injection location in the Greenidge boiler. For optimum AR-Lean performance, the CO concentration at the point of N-agent/OFA injection should be below 5000 ppm. The Greenidge boiler is characterized by upper furnace fluctuations in gas concentrations, and contains zones that have simultaneously high levels of CO and O 2 due to incomplete mixing. To simulate boiler design, two cooling arrays were installed in the furnace of the BSF: one simulating the high temperature secondary superheater and one simulating the reheater. The pilot-scale test results demonstrated that pulsations of CO and O 2 concentrations did not affect the performance of basic reburning, but decreased NO x reduction of SNCR by about 10% for tested experimental configuration. Performance in combined reburning + SNCR tests was almost independent on pulsing frequency and the reburning fuel flow rate, but decreased with pulsing amplitude. Results demonstrated that about 70-80% NO reduction could be achieved under Greenidge conditions using an optimized reburning + SNCR regime. Another objective of Task 2.4 was to evaluate coal as a reburning fuel. The results of the experiments indicated that the four tested bituminous coals were capable of providing reasonably high NO x control in basic reburning at the conditions available at the full-scale boilers. Over 90% NO x reduction could be achieved in AR with utilization of coal as a reburning fuel. The most effective variant of AR was reburning + SNCR followed by AR-Lean and AR-Rich. Tests showed that injection of promoters could significantly improve the efficiency of AR. Proof-of-concept tests in a Btu/hr combustion facility in Task 2.5 provided a final indication of the viability of the AR technology. The performance goals of Phase II to reduce NO x by up to 95% with net emissions less than 0.06 lb/10 6 Btu and to minimize other pollutants (N 2 O and NH 3 ) to levels lower than reburning and SNCR have been met. The following conclusions were drawn from experimental data obtained in different combustion facilities ranging in firing rate from to Btu/hr: AR provides up to 95% NO x reduction. AR-Lean + SNCR and Reburning + SNCR are the most effective AR configurations, followed by AR-Lean and AR-Rich. Promoters can increase the efficiency of NO x reduction in AR. Promoters are most effective at a small amount of the reburning fuel (6-10% of total fuel heat input). This provides the means to improve NO x reduction and simultaneously decrease the amount of reburning fuel required, relative to basic reburning. xxi

23 g DOE Contract No. DE-AC22-95PC95251 Final Report In Task 2.6 economic and market analyses of AR technologies were updated. The main driver to implement AR is NO x controls required in ozone non-attainment areas or areas which transport pollutants into ozone non-attainment areas. In the Northeastern portion of the country, this thirtyseven-state region consists of Pennsylvania and the States North and East of that state. This region can potentially be expanded to include Texas and all states North and East of this state. The NO x control requirements developed by the EPA to date have been based on attaining the current National Ambient Air Quality Standards (NAAQS). However, the EPA has issued revised NAAQS for ozone and fine particulate that are substantially lower than the current standards. Since NO x is a precursor of both pollutants, achieving the new NAAQS will require even greater reductions in NO x emissions which provides additional driver for AR technologies. The size of the market for AR technologies has been estimated to be above $110 million by considering the existing and projected CAAA regulations, the power plants affected by the regulations, and industry projections for the mix of NO x control technologies necessary for cost effective compliance with these regulations. Economic analysis demonstrates a considerable economic advantage of AR technologies in comparison with existing commercial NO x control techniques, such as basic reburning, SNCR, and SCR. Particularly for deep NO x control, coal-based AR technologies are 50% less expansive than SCR for the same level of NO x control. All project objectives and technical performance goals have been met or exceeded, and it was demonstrated that AR technologies could achieve high efficiency and low cost NO x control. xxii

24 g DOE Contract No. DE-AC22-95PC95251 Final Report 1.0 Introduction This project develops a family of novel Second Generation Advanced Reburning (SGAR) NO x control technologies, which have the potential to achieve 95% NO x control in coal fired boilers at a significantly lower cost than Selective Catalytic Reduction (SCR). AR systems integrate basic reburning and injection of N-agents (nitrogen-containing compounds capable of reducing NO, typically ammonia or urea). The AR systems are intended for EPA SIP Call compliance that requires to reduce NOx emissions from coal-fired facilities to the level of 0.15 lb/mmbtu in 22 states. Specific features of the new AR systems in comparison with basic reburning include: Introduction of reburning fuel representing a small portion of the total fuel heat input, to provide slightly fuel-rich conditions in the reburning zone. N-agent injection at one or two locations, which may include the reburning zone, the point of overfire air injection, and/or downstream of overfire air injection. Injection of promoter additives which enhance the effectiveness of the N-agent. The project was conducted in two phases over a five year period. The work included a combination of analytical and experimental studies to confirm the process mechanisms, identify optimum process configurations, and develop a design methodology for full-scale applications. Phase I was conducted from October, 1995 to September, 1997 and included both analytical studies and tests in bench and pilot scale test rigs. Phase II is conducted over a 45 month period (October, 1997 June, 2001). Phase II is based on the Phase I results and includes evaluation of alternate promoters, development of alternative reburning fuel and N-Agent jet mixing systems, and scale-up. This report consists of 13 Sections and 6 Appendices. Sections 1 through 4 describe background of the AR technology, summary of Phase I results and Phase II objectives. A detailed description of Phase I results is presented in Appendix A. Sections 5 through 11 describe Phase II results. Sections 12 and 13 present conclusions and referenced literature. Appendices B through F include chemical mechanisms, thermodynamic property data used in kinetic modeling, and a description of Computational Fluid Dynamic (CFD) modeling used to correlate the droplet size of liquid N-agent with its evaporation time. Extensive nomenclature is used in the description of different aspects of AR throughout this report. Figure 1-1 shows sketch of a boiler and summarizes the nomenclature for the various regions of the AR process. The region upstream of the reburning fuel injection is referred to as the primary zone or the main combustion zone. Combustion in the main combustion zone occurs in fuel lean environment so that the primary zone Stoichiometric Ratio (SR 1 ) is greater than 1.0. The initial NO concentration in this zone is referred to as NO i. The region between the reburning fuel and overfire air (OFA) injection is referred to as the reburning zone and is maintained at stoichiometry SR 2 which is usually less than 1.0. In AR-Rich N-agent and promoter are injected into the reburning zone, typically with a delay after reburning fuel injection. OFA is injected to complete combustion, downstream at lower furnace gas temperatures (which drop rapidly due to the heat exchange surfaces used for steam generation). In AR-Lean, N-agent and promoter are injected along with OFA. The downstream region of OFA injection referred to as the burnout zone. Typically OFA serves as the carrier gas for injecting an N-agent and promoter in AR-Lean. This zone is always fuel lean, at a stoichiometric ratio (SR 3 ) greater than 1.0. An N-agent can also be injected (with or without promoter) downstream of the OFA injection location into the burnout zone at furnace gas 1-1

25 g DOE Contract No. DE-AC22-95PC95251 Final Report conditions (particularly temperature) characteristic of the Selective Non-Catalytic Reduction (SNCR) process. This variant is called reburning + SNCR. N-agent + promoter Reburning + SNCR AR-Lean AR-Rich Overfire air (OFA) Reburning Fuel Burnout Zone Reburning Zone Air Main Fuel Primar Combustion Zone Figure 1-1. Schematic of different variants of AR. 1-2

26 g DOE Contract No. DE-AC22-95PC95251 Final Report 2.0 Background 2.1 High Ef ficiency NO x Control under Title 1 of the CAAA Title 1 of the Clean Air Act Amendment (CAAA) of 1990 requires NO x controls in ozone nonattainment areas. The initial Title 1 regulations required Reasonably Available Control Technologies (RACT). In most areas, the NO x levels for RACT were based on Low NO x Burners (LNB) and were in the range of 0.4 to 0.5 lb/10 6 Btu. As a result, there has been little industry demand for higher efficiency and more expensive NO x controls such as reburning, SNCR, and SCR. Over the last ten years, U.S. Environmental Protection Agency (EPA) has developed most of the specific NO x regulations authorized by the CAAA. The most stringent NO x controls are required in ozone non-attainment areas or areas which transport pollutants into ozone non-attainment areas. In the Northeast, EPA has defined the Northeast Ozone Transport Region (NEOTR) consisting of Pennsylvania and the states North and East. In that zone, NO x reductions of up to 75% are required by 2003 (SIP Call) with the potential for even deeper controls. The new control levels correspond to an average utility boiler NO x emission rate of 0.15 lb/10 6 Btu. EPA is now considering expanding the NEOTR to include Texas and all states North and East. In this 37 state region, it is projected that NO x emissions may need to be reduced by as much as 85%. As these specific regulations have developed, the trend has been towards cost effective emission controls. Rather than setting specific limits for each plant, in many areas the regulations have been established to provide the flexibility to over-control on some units and under-control on others, if that approach is cost effective. This can be of considerable advantage since the cost of NO x control for some units (particularly smaller units) may be much higher than for others, on a basis of $/ton of NO x removed. This bubbling approach depends on the availability of NO x control technologies which can achieve NO x reductions greater than the nominal control levels (75-85%) at low cost. Therefore, the goal established by DOE for this project, 95% NO x control down to 0.06 lb/10 6 Btu, is appropriate. NO x control technologies which meet this goal will only be employed if their costs are competitive with conventional controls on a $/ton basis. At present, the only commercial NO x control technology capable of achieving such deep NO x control is SCR. With SCR, NO x is reduced to N 2 by reactions with N-agents on the surface of a catalyst. The SCR process effectively uses the N-agent. Injection at a Nitrogen Stoichiometric Ratio, NSR (NSR is defined as molar ratio of N atoms in N-agent to that in NO x ) of 1.0 typically achieves about 80% NO x reduction (i.e., 80% N- agent utilization). SCR is fully commercialized in Europe and Japan and there are several U.S. installations. This is the reason for its extensive use as the basis of NO x control requirements for post-ract. Since the post-ract NO x control requirements are largely based on SCR, achieving the required NO x levels with SCR is relatively easy. However, SCR is far from an ideal utility solution. There are several important problems, and cost leads the list. SCR requires a catalyst in the flue gas exhaust stream. This catalyst, and the associated installation and boiler modifications, are expensive. As SCR technology has advanced over the last decade, the cost has decreased; however, at present, the initial cost of an 80% NO x control SCR system for a coal fired boiler is still about a factor of four greater than that of LNB. Increasing the NO x control to 95% approximately doubles the SCR system cost. In addition, the SCR catalyst life is limited. Catalyst deactivation, through a number of mechanisms, typically limits catalyst life to about 4 years for coal fired applications. SCR catalysts are also toxic, 2-1